It’s a nice list, but I think the core point strikes me as liable to be simply false. I forget who it was presenting this evidence—it might even have been James Miller, it was someone at the Winter Intelligence conference at FHI—but they looked at (1) the economic gains to countries with higher average IQ, (2) the average gains to individuals with higher IQ, and concluded that (3) people with high IQ create vast amounts of positive externality, much more than they capture as individuals, probably mostly in the form of countries with less stupid economic policies.
Maybe if we’re literally talking about a pure speed and LTM pill that doesn’t affect at all, say, capacity to keep things in short-term memory or the ability to maintain complex abstractions in working memory, i.e., a literal speed and disk space pill rather than an IQ pill.
Absolutely—IQ is very important, especially in aggregate. And yet, I’d still bet that the next day people will just be moving faster.
I think its worth making the distinction between having hardware which can support complex abstractions and actually having good decision making software in there. Although it’d be foolish to ignore the former because it tends to lead to the latter, it seems to be the latter that is more directly important.
That, and the fact that people can generally support better software than they pick up on their own is what makes our goal here doable.
If this is true, it would affect my decisions about whether and how to have children. So I’d really like to see the source if you can figure out what it was.
They found that intelligence made a difference in gross domestic product. For each one-point increase in a country’s average IQ, the per capita GDP was $229 higher. It made an even bigger difference if the smartest 5 percent of the population got smarter; for every additional IQ point in that group, a country’s per capita GDP was $468 higher.
Citing “Cognitive Capitalism: The impact of ability, mediated through science and economic freedom, on wealth”. (PDF not immediately available in Google.)
Economic models of the loss caused by small intelligence decrements due to lead in drinking water predict significant effects of even a few points decrease (Salkever 1995; Muir and Zegarac 2001). Because the models are roughly linear for small changes, they can be inverted to estimate societal effects of improved cognition. The Salkever model estimates the increase in income due to one more IQ point to be 2.1% for men and 3.6% for women. (Herrnstein and Murray 1994) estimate that a 3% increase in overall IQ would reduce the poverty rate by 25%, males in jail by 25%, high-school dropouts by 28%, parentless children by 20%, welfare recipients by 18%, and out-of-wedlock births by 25%.
How important is intelligence to financial success? Using the NLSY79, which tracks a large group of young U.S. baby boomers, this research shows that each point increase in IQ test scores raises income by between $234 and $616 per year after holding a variety of factors constant. Regression results suggest no statistically distinguishable relationship between IQ scores and wealth. Financial distress, such as problems paying bills, going bankrupt or reaching credit card limits, is related to IQ scores not linearly but instead in a quadratic relationship. This means higher IQ scores sometimes increase the probability of being in financial difficulty.
One could also phrase this as: “if we control for factors which we know to because by intelligence, such as highest level of education, then mirabile dictu! intelligence no longer increases income or wealth very much!”; or, “regressions are hard, let’s go shopping.”
In the XXIst century within wealthy countries, people work hard primarily to gain social status. We often make the mistake of tying up wealth with social status, but most of the wealthy people we admire are also consumed by their great jobs. Celine Dion is very wealthy, yet she would still give one show every single day, including week-ends. I think most professors would feel exploited if they had to lecture every single day. Bill Gates is very wealthy and universally admired, however, as we may expect, he worked nights and week-ends as chairman of Microsoft. Every year he would read 100 papers from Microsoft employees about the state of the company.
...For many, wealth is merely a stepping stone to intense work. This may explain why people with higher IQs are not wealthier (Zagorsky, 2008): high IQ people may have an easier time getting rewarding work so they need less wealth....I used to openly worry that robots would steal our jobs and leave most of us in poverty. I have now concluded that I was underestimating the pull of prestige among human beings. We will make up jobs out of thin air if we need to.
I show that in a conventional Ramsey model, between one-fourth and one-half of the global income distribution can be explained by a single factor: The effect of large, persistent differences in national average IQ on the private marginal product of labor. Thus, differences in national average IQ may be a driving force behind global income inequality. These persistent differences in cognitive ability—which are well-supported in the psychology literature—are likely to be somewhat malleable through better health care, better education, and especially better nutrition in the world’s poorest countries. A simple calibration exercise in the spirit of Bils and Klenow (2000) and Castro (2005) is conducted. I show that an IQ-augmented Ramsey model can explain more than half of the empirical relationship between national average IQ and GDP per worker. I provide evidence that little of the IQ-productivity relationship is likely to be due to reverse causality.
One question of interest is whether the IQ-productivity relationship has strengthened or weakened over the past few decades. Shocks such as the Great Depression and the Second World War were likely to move nations away from their steady-state paths. Further, many countries have embraced market economies in recent decades, a policy change which is likely to have removed non-IQ-related barriers to riches.11 Accordingly, one would expect the IQ-productivity relationship to have strengthened over the decades.
As Table 2 shows, I indeed found this to be the case. I used LV’s IQ data along with Penn World Table data for each decade from 1960 through 1990 (1950 only had 38 relevant observations, and so is omitted). As before, equation (3) was used to estimate the IQ-productivity relationship, while the IQ-elasticity of wages is assumed to equal 1 for simplicity. Both the unconditional R2 and the fraction of the variance explained by the IQ-wage relationship increase steadily across the decades. This is true regardless of the capital share parameter in question. Further, the log-slope of the IQ-productivity relationship has also increased.
11: Lynn and Vanhanen (2002) hypothesize that national average IQ and market institutions are the two crucial determinants of GDP per capita. They provide some bivariate regressions supporting this hypothesis; they show that both variables together explain much more—about 75% of the variance in the level of GDP per capita—than either variable alone, each of which can explain roughly 50%.
The Ramsey-style model of Manuelli and Seshadri (2005) would be a natural extension: In their model, ex-ante differences in total factor productivity of at most 27% interact with education decisions and fertility choices to completely replicate the span of the current global income distribution. In their calibration—less naïve and more complex then the one I present—a 1% rise in TFP (e.g., 1 IQ point) causes a 9% rise in steady- state productivity. Manuelli and Seshadri leave unanswered the question of what those ex-ante differences in TFP might be, but persistent differences in national average IQ are a natural candidate.
Only the first paragraph is wrong (mixed it up with a paper on the Swiss iodization experience I’m using in a big writeup on iodide self-experimentation). Fixed.
We assumed the change in cognitive ability resulting from declines in BLLs, on the basis of published meta-analyses, to be between 0.185 and 0.323 IQ points for each 1 g/dL blood lead concentration. These calculations imply that, because of falling BLLs, U.S. preschool-aged children in the late 1990s had IQs that were, on average, 2.2-4.7 points higher than they would have been if they had the blood lead distribution observed among U.S. preschool-aged children in the late 1970s. We estimated that each IQ point raises worker productivity 1.76-2.38%. With discounted lifetime earnings of $723,300 for each 2-year-old in 2000 dollars, the estimated economic benefit for each year’s cohort of 3.8 million 2-year-old children ranges from $110 billion to $319 billion.
...We calculated the economic benefit realized by reduced lead exposure in the United States since the late 1970s through a series of steps, each associated with a component of the model in Figure 1. First, we estimated the amount by which BLLs have fallen over time through secondary analysis of data from the National Health and Nutrition Examination Surveys (NHANES). Second, we applied estimates from published studies of the strength, shape, and magnitude of the association between BLLs and cognitive ability test scores. In particular, we examined two published meta-analyses to arrive at estimates of the ratio of change in BLL to change in IQ. Third, on the basis of a brief review of literature on the association between cognitive ability and earning potential, we estimated the percentage change in earnings associated with absolute differences in IQ levels. Fourth, we calculated the present value (2000 dollars) of the percentage change in earnings.
...Schwartz (6) calculated that the total effect of a 1-point difference in cognitive ability is a 1.76% difference in earnings. Of this amount, 0.5% is the direct effect of ability on earnings. Schwartz (6) took this estimate from an econometric study by Griliches (19) that was representative of other econometric studies from the 1970s. Schwartz (6) assumed that a given difference in IQ scores observed in school-aged children can be expected to lead to a comparable difference in achieved cognitive ability in young adults.
The indirect effect of ability on earnings, which accounts for the remaining 1.26% difference, is modeled through two pathways. One is the effect of ability on years of schooling multiplied by the effect of years of schooling on hourly earnings. Needleman et al. (20) reported that a 4.5-point difference in IQ between groups with high tooth lead and with low tooth lead was associated with a 0.59 difference in grade level attained. The ratio of the two numbers implies a difference of 0.131 years of schooling for 1 IQ point. If each additional year of schooling results in a 6% increase in hourly wages, 1 IQ point would lead to a 0.79% increase in expected earnings through years of education. Second, Schwartz (6) modeled ability as influencing employment participation through influence on high school graduation. On the basis of the analysis of Needleman et al. (20) and 1978 survey data reported by Krupnick and Cropper (21), Schwartz (6) calculated that 1 point in IQ is associated with a 4.5% difference in probability of graduating from high school and that high school graduation is associated with a 10.5% difference in labor force participation. On the assumption of an equivalent percentage change in annual earnings, this leads to a 0.47% difference in expected earnings. Salkever (22) published an alternate estimation of the effect of cognitive ability on earnings. Salkever directly estimated the effect of ability on annual earnings, among those with earnings. The estimated association of ability with annual earnings incorporates both the effect of ability on hourly earnings and its effect on annual hours of work. He also added a direct pathway from ability to work participation independent of education.
According to Salkever (22), a 1-point difference in ability is associated with a 1.931% difference in earnings for males and a 3.225% difference for females. The direct effect on earnings is 1.24% for males and 1.40% for females. Salkever (22) analyzed income and educational attainment data from the 1990 wave of the National Longitudinal Study of Youth (NLSY) in combination with AFQT scores collected during 1979–1980, when the respondents were 14–23 years of age.
For the indirect effect of ability on schooling attainment, Salkever (22) reported that a 1-point difference was associated with 0.1007 years of schooling attained for both males and females in the NLSY data. Also, 1 year of schooling attainment raised hourly earnings by 4.88% for males and 10.08% for females in the 1990 NLSY data. According to these results, a 1-point difference in ability is associated, through an indirect effect on schooling, with a 0.49% difference in earnings for males and a 1.10% difference in earnings for females.
Salkever (22) reported that the direct effect of a 1-point difference in ability was a 0.1602 percentage point difference in probability of labor force participation for males and a 0.3679 percentage point difference for females. In addition, he calculated that 1 year of schooling raised labor force participation rates by 0.3536 percentage points for males and 2.8247 percentage points for females. Subtracting the other components from the totals, a 1-point change in cognitive ability is associated with a difference in earnings of 0.20% for males and 0.72% for females through effects on labor force participation. Finally, in an analysis of the 1990 NLSY earnings data, Neal and Johnson (23) reported smaller estimates of the effect of cognitive ability on earnings. They included workers who took the AFQT test when they were 14–18 years of age and excluded those who took the AFQT test at 19–23 years of age to make the test scores more comparable. They also estimated the total effect of ability on hourly earnings by excluding schooling variables. Their estimates indicate that a 1point difference in AFQT scores is associated with a 1.15% difference in earnings for men and a 1.52% difference for women. Their estimate of the direct effect of ability on hourly earnings, controlling for schooling, is 0.83% for men; they reported no estimate for women.
The analysis of Neal and Johnson (23) has no link from ability to labor force participation. According to Salkever (22), a 1-point difference in ability leads to a 0.20% difference for males and 0.72% for females. If we add Salkever’s figures (22) to the estimates from Neal and Johnson (23), the total effect of a 1-point difference in ability on earnings is 1.35% for males and 2.24% for females.
Their summary estimate from pg5/567 is a lower-middle-upperbound of each IQ point is worth, in net present value 2000 dollars: 12,700-14,500-17,200.
(Note that these figures, as usual, are net estimates of the value to an individual: so they are including zero-sum games and positional benefits. They aren’t giving estimates of the positive externalities or marginal benefits.)
We analyze the effect of the average level of intelligence on different measures of the quality of institutions, using a 2006 cross-sectional sample of 113 countries. The results show that average IQ positively affects all the measures of institutional quality considered in our study, namely government eciency, regulatory quality, rule of law, political stability and voice and accountability. The positive effect of intelligence is robust to controlling for other determinants of institutional quality.
I used data from the NLSY79 which is an ongoing longitudinal study that follows the lives of a large sample of Americans born in 1957-64. Specifically, I used the nationally representative subsample comprising more than 6000 individuals...The unstandardized slope coefficient is 0.025 (95% CI: 0.023-0.027). Because the dependent variable is logarithmic, this coefficient, when multiplied by 100, can be (approximately) interpreted as the percent change in income in (unlogged) dollars associated with a 1 IQ point change.[Note] Therefore, one additional IQ point predicts a 2.5% boost in income. The standardized effect size, or correlation, is 0.36 and the R squared is 13%.
A meta-study of repeated prisoner’s dilemma experiments run at numerous universities suggests that students cooperate 5% to 8% more often for every 100 point increase in the school’s average SAT score.
This finding was the first of its kind: In prisoner’s dilemmas, smarter groups really were more cooperative. Since then other researchers have found similar results, some of which I discuss in Section III of this article for the Asian Development Review. It looks like intelligence is a form of social intelligence...Does that happen in the real world? If it does, does it mean that there are negative political externalities to low-skill immigration? That’s a topic for a later time. Another worthy question: Why would high IQ groups be more cooperative anyway? Isn’t cynicism intelligent? Sure, sometimes, but the political entrepreneur who can find a way to sustain a truce can probably skim quite a lot of the resulting prosperity off for herself. And people who are better at solving the puzzles in an IQ test are probably better at solving the puzzles of human interaction.
We show that a country’s average IQ score is a useful predictor of the wages that immigrants from that country earn in the U.S., whether or not one adjusts for immigrant education. Just as in numerous microeconomic studies, 1 IQ point predicts 1% higher wages, suggesting that IQ tests capture an important difference in cross-country worker productivity. In a cross-country development accounting exercise, about one-sixth of the global inequality in log income can be explained by the effect of large, persistent differences in national average IQ on the private marginal product of labor. Taken together with the results of Jones and Schneider (2006) and Hanushek and Kimko (2000), this suggests that cognitive skills matter more for groups than for individuals.
Background: Results from previous studies show that the cognitive ability of off spring might be irreversibly damaged as a result of their mother’s mild iodine deficiency during pregnancy. A reduced intelligence quotient (IQ) score has broad economic and societal cost implications because intelligence affects wellbeing, income, and education outcomes. Although pregnancy and lactation lead to increased iodine needs, no UK recommendations for iodine supplementation have been issued to pregnant women. We aimed to investigate the cost-effectiveness of iodine supplementation versus no supplementation for pregnant women in a mildly to moderately iodine-deficient population for which a population- based iodine supplementation programme-for example, universal salt iodisation-did not exist.
Methods: We systematically searched MEDLINE, Embase, EconLit, and NHS EED for economic studies that linked IQ and income published in all languages until Aug 21, 2014. We took clinical data relating to iodine deficiency in pregnant women and the effect on IQ in their children aged 8-9 years from primary research. A decision tree was developed to compare the treatment strategies of iodine supplementation in tablet form with no iodine supplementation for pregnant women in the UK. Analyses were done from a health service perspective (analysis 1; taking direct health service costs into account) and societal perspective (analysis 2; taking education costs and the value of an IQ point itself into account), and presented in terms of cost (in sterling, relevant to 2013) per IQ point gained in the off spring. We made data-supported assumptions to complete these analyses, but used a conservative approach that limited the benefits of iodine supplementation and overestimated its potential harms.
Findings: Our systematic search identified 1361 published articles, of which eight were assessed to calculate the monetary value of an IQ point. A discounted lifetime value of an additional IQ point based on earnings was estimated to be £3297 (study estimates range from £1319 to £11 967) for the off spring cohort. Iodine supplementation was cost saving from both a health service perspective (saving £199 per pregnant woman [sensitivity analysis range -£42 to £229]) and societal perspective (saving £4476 per pregnant woman [sensitivity analysis range £540 to £4495]), with a net gain of 1·22 IQ points in each analysis. Base case results were robust to sensitivity analyses.
Interpretation: Iodine supplementation for pregnant women in the UK is potentially cost saving. This finding also has implications for the 1·88 billion people in the 32 countries with iodine deficiency worldwide. Valuation of IQ points should consider non-earnings benefits-eg, health benefits associated with a higher IQ not germane to earnings.
IQ estimates:
Our systematic search identified 1361 published articles, of which eight studies 47-54 passed quality criteria and were assessed to calculate the monetary value of an IQ point (appendix p 4). The quality criteria were as follows: an individual’s IQ is used and is not a proxy; variables are clearly specified; IQ measure follows a conventional normal distribution with a mean of 100 and standard deviation of 15 or sufficient information is included in the study to allow the IQ measure’s distribution to be converted into one (for cross study comparability); and the results reported in currency form have the applicable year stated. Most of the studies valued an IQ point on the basis of its effect on an individual’s income (appendix p 3). The issue of differences in scaling of IQ tests hindered the comparability across studies. The value of an IQ point, derived from the systematic search and applied to the unborn cohort, comes from the lifetime earnings premium of an additional IQ point. This is calculated to be £3297 (study estimates range from £1319 to £11967; after adjustment with life tables).
One study looked at people’s willingness to pay (WTP) for an additional IQ point. 4 Five studies used econometric regressions to determine the individuals IQ’s effect on their subsequent income, 5-9 whereas two studies were cost benefit analysis on reducing lead exposure. 10,11 Only one of the studies included in the systematic literature search was not set in the USA. 5...In keeping with the conservative nature of the model, the relatively high earnings premium from IQ points from Schwartz 10 and Salkever 11 are excluded on the basis that the effect may be overstated.
The 8 studies are listed on pg8 of the appendix, Table 1:
(Note that by including covariates that are obviously caused by IQ rather than independent, and excluding any attempt at measuring the many positive externalities of greater intelligence, these numbers can usually be considered substantial underestimates of country-wide benefits.)
This report uses recent economic modelling to relate cognitive skills – as measured by PISA and other international instruments – to economic growth. The relationship indicates that relatively small improvements in the skills of a nation’s labour force can have very large impacts on future well-being.
...A modest goal of having all OECD countries boost their average PISA scores by 25 points over the next 20 years – which is less than the most rapidly improving education system in the OECD, Poland, achieved between 2000 and 2006 alone – implies an aggregate gain of OECD GDP of USD 115 trillion over the lifetime of the generation born in 2010 (as evaluated at the start of reform in terms of real present value of future improvements in GDP) (Figure 1). Bringing all countries up to the average performance of Finland, OECD’s best performing education system in PISA, would result in gains in the order of USD 260 trillion (Figure 4). The report also shows that it is the quality of learning outcomes, not the length of schooling, which makes the difference. Other aggressive goals, such as bringing all students to a level of minimal proficiency for the OECD (i.e. reaching a PISA score of 400), would imply aggregate GDP increases of close to USD 200 trillion according to historical growth relationships (Figure 2).
...Using data from international student achievement tests, Hanushek and Kimko (2000) demonstrate a statistically and economically significant positive effect of cognitive skills on economic growth in 1960-90. Their estimates suggest that one country-level standard deviation higher test performance would yield around one percentage point higher annual growth rates. The country-level standard deviation is equivalent to 47 test-score points in the PISA 2000 mathematics assessment. Again, in terms of the PISA 2000 mathematics scores, 47 points would be roughly the average difference between Sweden and Japan (the best performer among OECD countries in 2000) or between the average Greek student and the OECD average score. One percentage point difference in growth is itself a very large value, because the average annual growth of OECD countries has been roughly 1.5%.
Their estimate stems from a statistical model that relates annual growth rates of real GDP per capita to the measure of cognitive skills, years of schooling, the initial level of income and a wide variety of other variables that might affect growth including in different specifications the population growth rates, political measures, or openness of the economies.
...The relationship between cognitive skills and economic growth has now been demonstrated in a range of studies. As reviewed in Hanushek and Woessmann (2008), these studies employ measures of cognitive skills that draw upon the international testing of PISA and of TIMSS (Trends in International Mathematics and Science Study) (along with earlier versions of these).7 The uniform result is that the international achievement measures provide an accurate measure of the skills of the labour force in different countries and that these skills are closely tied to economic outcomes.8
...While the PISA tests are now well-known throughout the OECD, the history of testing is less understood. Between 1964 and 2003, 12 different international tests of mathematics, science, or reading were administered to a voluntarily participating group of countries (see Annex Tables A1 and A2). These include 36 different possible scores for year-age-test combinations (e.g. science for students of grade 8 in 1972 as part of the First International Science Study or mathematics of 15-year-olds in 2000 as a part of the Programme on International Student Assessment). Only the United States participated in all possible tests. The assessments are designed to identify a common set of expected skills, which were then tested in the local language. It is easier to do this in mathematics and science than in reading, and a majority of the international testing has focused on mathematics and science. Each test is newly constructed, until recently with no effort to link to any of the other tests. While the analysis here focuses on mathematics and science, these scores are highly correlated with reading test scores and employing just mathematics and science performance does not distort the growth relationship that is estimated; see Hanushek and Woessmann (2009). The goal here is construction of consistent measures at the national level that will allow comparing performance across countries, even when they did not each participate in a common assessment.
...The simplest overview of the relationship is found in Figure 6 that plots regional growth in real per capita GDP between 1960 and 2000 against average test scores after allowing for differences in initial GDP per capita in 1960.14 Regional annual growth rates, which vary from 1.4% in Sub-Saharan Africa to 4.5% in East Asia, fall on a straight line.15 But school attainment, when added to this regression, is unrelated to growth- rate differences. Figure 6 suggests that, conditional on initial income levels, regional growth over the last four decades is completely described by differences in cognitive skills.
Second, to tackle the most obvious reverse-causality issues, Hanushek and Woessmann (2009) separate the timing of the analysis by estimating the effect of scores on tests conducted until the early 1980s on economic growth in 1980-2000. In this analysis, available for a smaller sample of countries only, test scores pre-date the growth period. The estimate shows a significant positive effect that is about twice as large as the coefficient used in the simulations here.
Needless to say, “cognitive skills” here is essentially an euphemism for intelligence/IQ.
A modest goal of having all OECD countries boost their average PISA scores by 25 points over the next 20 years – which is less than the most rapidly improving education system in the OECD, Poland, achieved between 2000 and 2006 alone – implies an aggregate gain of OECD GDP of USD 115 trillion over the lifetime of the generation born in 2010 (as evaluated at the start of reform in terms of real present value of future improvements in GDP) (Figure 1). Bringing all countries up to the average performance of Finland, OECD’s best performing education system in PISA, would result in gains in the order of USD 260 trillion (Figure 4). The report also shows that it is the quality of learning outcomes, not the length of schooling, which makes the difference. Other aggressive goals, such as bringing all students to a level of minimal proficiency for the OECD (i.e. reaching a PISA score of 400), would imply aggregate GDP increases of close to USD 200 trillion according to historical growth relationships (Figure 2).
Education and general intelligence both serve to inform opinions, but do they lead to greater attitude extremity? We use questions on economic policy, social issues, and environmental issues from the General Social Survey to test the impact of education and intelligence on attitude extremity, as measured by deviation from centrist or neutral positions. Using quantile regression modeling, we find that intelligence is a moderating force across the entire distribution in economic, social, and environmental policy beliefs. Completing high school strongly correlates to reduced extremity, particularly in the upper quantiles. College education increases attitude extremity in the lower tail of environmental beliefs. The relevance of the low extremity tail (lower quantiles) to potential swing-voters and the high extremity tail (upper quantiles) to a political party’s core are discussed.
The authors analysed data from the 2007 Adult Psychiatric Morbidity Survey in England. The participants were adults aged 16 years or over, living in private households in 2007. Data from 6870 participants were included in the study...Happiness is significantly associated with IQ. Those in the lowest IQ range (70–99) reported the lowest levels of happiness compared with the highest IQ group (120–129). Mediation analysis using the continuous IQ variable found dependency in activities of daily living, income, health and neurotic symptoms were strong mediators of the relationship, as they reduced the association between happiness and IQ by 50 %
I think that you might be confusing causation and correlation here. Countries that started to industrialize earlier have higher average IQ and higher GDP per capita. That would produce the effect you refer to. Whether or not the increased intelligence then contributes to further economic growth is a different matter.
What third factor producing both higher IQ and then industrialization are you suggesting?
Obviously you’re not suggesting anything as silly as the industrialization causes all observed IQ changes, because that simply doesn’t explain all examples, like East Asian countries:
A crucial question is whether IQ differences across countries are a simple case of reverse causation: Do high-income countries simply develop higher IQ’s? We address this question in a number of ways, but the most important is likely to be this simple fact: East Asian countries had high average IQ’s—at or above the European and U.S. averages—well before they entered the ranks of the high income countries. This is precisely the opposite of what one would expect if the IQ-productivity relationship were merely epiphenomenal.
East Asian countries had high average IQ’s—at or above the European and U.S. averages—well before they entered the ranks of the high income countries.
That suggests that the correlation would have been less at that earlier time, which suggests the idea that the correlation of average IQ and average income has varied over history. Perhaps it has become stronger with increasing technological level—that is, more opportunities to apply smarts?
That certainly seems possible. Imagine a would-be programming genius who is born now, versus born in the Stone Age—he could become the wealthiest human to ever live (Bill Gates) or just the best hunter in the tribe (to be optimistic...).
Based on their pioneering research two research questions were developed: does intelligence lead to wealth or does wealth lead to intelligence or are other determinants involved? If a nation’s intelligence increases wealth, how does intelligence achieve this? To answer them we need longitudinal studies and theoretical attempts, investigating cognitive ability effects at the levels of individuals, institutions and societies and examining factors which lie between intelligence and growth. Two studies, using a cross-lagged panel design or latent variables and measuring economic liberty, shares of intellectual classes and indicators of scientific-technological accomplishment, show that cognitive ability leads to higher wealth and that for this process the achievement of high ability groups is important, stimulating growth through scientific-technological progress and by influencing the quality of economic institutions.
...cognitive skills—intelligence quotient scores, math skills, and the like—have only a modest influence on individual wages, but are strongly correlated with national outcomes. Is this largely due to human capital spillovers? This paper argues that the answer is yes. It presents four different channels through which intelligence may matter more for nations than for individuals: (i) intelligence is associated with patience and hence higher savings rates; (ii) intelligence causes cooperation; (iii) higher group intelligence opens the door to using fragile, high-value production technologies; and (iv) intelligence is associated with supporting market-oriented policies.
Plots of mean IQ and per capita real Gross Domestic Product for groups of 81 and 185 nations, as collected by Lynn and Vanhanen, are best fitted by an exponential function of the form: GDP = a 10^b\(IQ), where a and b are empirical constants. Exponential fitting yields markedly higher correlation coefficients than either linear or quadratic. The implication of exponential fitting is that a given increment in IQ, anywhere along the IQ scale, results in a given percentage in GDP, rather than a given dollar increase as linear fitting would predict. As a rough rule of thumb, an increase of 10 points in mean IQ results in a doubling of the per capita GDP.
....In their book, IQ and the Wealth of Nations, Lynn and Vanhanen (2002) present a table listing for 81 nations the measured mean IQ and the per capita real Gross Domestic Product as of 1998 (their Table 7.7). They subsequently extend this to all 185 nations, using estimated IQs for the 104 new entries based chiefly on IQ values for immediate neighbors (their Table 8.9). In both cases they observe a significant correlation between IQ and GDP, with linear correlation factors R^2 = 0.537 for the 81-nation group and 0.389 for 185 nations. McDaniel and Whetzel have extended the examination of correlations to quadratic fitting in a paper that demonstrates the robustness of these correlations to minor variations in individual IQ values (McDaniel & Whetzel, in press). But an even stronger correlation is found if the fitting is exponential rather than linear or quadratic.
It peeves me when scatterplots of GDP per capita versus something else use a linear scale—do they actually think the difference between $30k and $20k is anywhere near as important as that between $11k and $1k? And yet hardly anybody uses logarithmic scales.
Likewise, the fit looks a lot less scary if you write it as ln(GDP) = A + B*IQ.
Yes, Dickerson does point out that his exponential fit is a linear relationship on a log scale. For example, he does show a log-scale in figure 3 (pg3), fitting the most reliable 83 nation-points on a plot of log(GDP) against mean IQ in which the exponential fit looks exactly like you would expect. (Is it per capita? As far as I can tell, he always means per capita GDP even if he writes just ‘GDP’.) Figure 4 does the same thing but expands the dataset to 185 nations. The latter plot should probably be ignored given that the expansion comes from basically guessing:
In their book, IQ and the Wealth of Nations, Lynn and Vanhanen (2002) present a table listing for 81 nations the measured mean IQ and the per capita real Gross Domestic Product as of 1998 (their Table 7.7). They subsequently extend this to all 185 nations, using estimated IQs for the 104 new entries based chiefly on IQ values for immediate neighbors (their Table 8.9).
I dunno. I’ve given it a try and while it’s easy enough to reproduce the exponential fit (and the generated regression line does fit the 81 nations very nicely), I think I screwed up somehow reproducing the smart fraction equation because the regression looks weird and trying out the smart-fraction function (using his specified constants) on specific IQs I don’t get the same results as in La Griffe’s table. And I can’t figure out what I’m doing wrong, my function looks like it’s doing the same thing as his. So I give up. Here is my code if you want to try to fix it:
lynn <- read.table(stdin(),header=TRUE,sep="")
Country IQ rGDPpc
Argentina 96 12013
Australia 98 22452
Austria 102 23166
Barbados 78 12001
Belgium 100 23223
Brazil 87 6625
Bulgaria 93 4809
Canada 97 23582
China 100 3105
Colombia 89 6006
Congo 65 822
Congo 73 995
Croatia 90 6749
Cuba 85 3967
CzechRepublic 97 12362
Denmark 98 24218
Ecuador 80 3003
Egypt 83 3041
EquatorialGuinea 59 1817
Ethiopia 63 574
Fiji 84 4231
Finland 97 20847
France 98 21175
Germany 102 22169
Ghana 71 1735
Greece 92 13943
Guatemala 79 3505
Guinea 66 1782
HongKong 107 20763
Hungary 99 10232
India 81 2077
Indonesia 89 2651
Iran 84 5121
Iraq 87 3197
Ireland 93 21482
Israel 94 17301
Italy 102 20585
Jamaica 72 3389
Japan 105 23257
Kenya 72 980
Lebanon 86 4326
Malaysia 92 8137
MarshallIslands 84 3000
Mexico 87 7704
Morocco 85 3305
Nepal 78 1157
Netherlands 102 22176
NewZealand 100 17288
Nigeria 67 795
Norway 98 26342
Peru 90 4282
Philippines 86 3555
Poland 99 7619
Portugal 95 14701
PuertoRico 84 8000
Qatar 78 20987
Romania 94 5648
Russia 96 6460
SierraLeone 64 458
Singapore 103 24210
Slovakia 96 9699
Slovenia 95 14293
SouthAfrica 72 8488
SouthKorea 106 13478
Spain 97 16212
Sudan 72 1394
Suriname 89 5161
Sweden 101 20659
Switzerland 101 25512
Taiwan 104 13000
Tanzania 72 480
Thailand 91 5456
Tonga 87 3000
Turkey 90 6422
UKingdom 100 20336
Uganda 73 1074
UnitedStates 98 29605
Uruguay 96 8623
WesternSamoa 87 3832
Zambia 77 719
Zimbabwe 66 2669
em <- lm(log(lynn$rGDPpc) ~ lynn$IQ); summary(em)
Call:
lm(formula = log(lynn$rGDPpc) ~ lynn$IQ)
Residuals:
Min 1Q Median 3Q Max
-1.6124 -0.3866 -0.0429 0.3363 2.0311
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 1.77760 0.51848 3.43 0.00097
lynn$IQ 0.07876 0.00583 13.52 < 2e-16
Residual standard error: 0.624 on 79 degrees of freedom
Multiple R-squared: 0.698, Adjusted R-squared: 0.694
F-statistic: 183 on 1 and 79 DF, p-value: <2e-16
# plot
plot (log(lynn$rGDPpc) ~ lynn$IQ)
abline(em)
# an attempt at La Griffe
erf <- function(x) 2 * pnorm(x * sqrt(2)) - 1
sf <- function(iq) ((69321/2) * (1 + erf(((iq - 108)/15) / sqrt(2))))
# check for sigmoid
# plot(c(85:130), sf(c(85:130)))
lg <- lm(log(lynn$rGDPpc) ~ sf(lynn$IQ)); summary(lg)
Call:
lm(formula = log(lynn$rGDPpc) ~ sf(lynn$IQ))
Residuals:
Min 1Q Median 3Q Max
-2.5788 -0.6857 0.0678 1.0521 1.5901
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 8.705620 0.126152 69.01 <2e-16
sf(lynn$IQ) 0.000121 0.000102 1.19 0.24
Residual standard error: 1.13 on 79 degrees of freedom
Multiple R-squared: 0.0175, Adjusted R-squared: 0.0051
F-statistic: 1.41 on 1 and 79 DF, p-value: 0.239
# same plotting code
(In retrospect, I’m not sure it’s even meaningful to try to fit the sf function with the constants already baked in, but since I apparently didn’t write it right, it doesn’t matter.
Hm, one thing I notice is that you look like you’re fitting sf against log(gdp). I managed to replicate his results in octave, and got a meaningful result plotting smart fraction against gdp.
My guess at how to change your code (noting that I don’t know R):
sf <- function(iq,c) ((69321/2) * (1 + erf(((iq - c)/15) / sqrt(2))))
lg <- lm(lynn$rGDPpc ~ sf(lynn$IQ,108)); summary(lg)
That should give you some measure of how good it fits, and you might be able to loop it to see how well the smart fraction does with various thresholds.
I can’t tell whether that works since you’re just using the same broken smart-fraction sf predictor; eg. sf(107,108) ~> 32818, while the first smart fraction page’s table gives a Hong Kong regression line of 19817 which is very different from 33k.
Hmmm. I agree that it doesn’t match. What if by ‘regression line’ he means the regression line put through the sf-gdp data?
That is, you should be able to calculate sf as a fraction with
sf <- function(iq,c) ((1/2) * (1 + erf((iq-c)/(15*sqrt(2)))))
And then regress that against gdp, which will give you the various coefficients, and a much more sensible graph. (You can compare those to the SFs he calculates in the refinement, but those are with verbal IQ, which might require finding that dataset / trusting his, and have a separate IQ0.)
Comparing the two graphs, I find it interesting that the eight outliers Griffe mentions (Qatar, South Africa, Barbados, China, and then the NE Asian countries) are much more noticeable on the SF graph than the log(GDP) graph, and that the log(GDP) graph compresses the variation of the high-income countries, and gets most of its variation from the low-income countries; the situation is reversed in the SF graph. Since both our IQ and GDP estimates are better in high-income countries, that seems like a desirable property to have.
With outliers included, I’m getting R=.79 for SF and R=.74 for log(gdp). (I think, I’m not sure I’m calculating those correctly.)
Trying to rederive the constants doesn’t help me, which is starting to make me wonder if he’s really using the table he provided or misstated an equation or something:
R> sf <- function(iq,f,c) ((c/2) * (1 + erf((iq-f)/(15*sqrt(2)))))
R> summary(nls(rGDPpc ~ sf(IQ,f,c), lynn, start=list(f=110,c=40000)))
Formula: rGDPpc ~ sf(IQ, f, c)
Parameters:
Estimate Std. Error t value Pr(>|t|)
f 99.64 3.07 32.44 < 2e-16
c 34779.17 6263.90 5.55 3.7e-07
Residual standard error: 5310 on 79 degrees of freedom
Number of iterations to convergence: 4
Achieved convergence tolerance: 8.22e-06
If you double 34779 you get very close to his $69,321 so there might be something going wrong due to the 1⁄2 that appears in uses of the erf to make a cumulative distribution function, but I don’t how a threshold of 99.64 IQ is even close to his 108!
(The weird start values were found via trial-and-error in trying to avoid R’s ‘singular gradient error’; it doesn’t appear to make a difference if you start with, say, f=90.)
Most importantly, we appear to have figured out the answer to my original question: no, it is not easy. :P
So, I started off by deleting the eight outliers to make lynn2. I got an adjusted R^2 of 0.8127 for the exponential fit, and 0.7777 for the fit with iq0=108.2.
My nls came back with an optimal iq0 of 110, which is closer to the 108 I was expecting; the adjusted R^2 only increases to 0.7783, which is a minimal improvement, and still slightly worse than the exponential fit.
The value of the smart fraction cutoff appears to have a huge impact on the mapping from smart fraction to gdp, but doesn’t appear to have a significant effect on the goodness of fit, which troubles me somewhat. I’m also surprised that deleting the outliers seems to have improved the performance of the exponential fit more than the smart fraction fit, which is not what I would have expected from the graphs. (Though, I haven’t calculated this with the outliers included in R, and I also excluded the Asian data, and there’s more fiddling I can do, but I’m happy with this for now.)
> sf <- function(iq,iq0) ((1+erf((iq-iq0)/(15*sqrt(2))))/2)
> egdp <- function(iq,iq0,m,b) (m*sf(iq,iq0)+b)
> summary(nls(rGDPpc ~ egdp(IQ,iq0,m,b), lynn2, start=list(iq0=110,m=40000,b=0)))
Formula: rGDPpc ~ egdp(IQ, iq0, m, b)
Parameters:
Estimate Std. Error t value Pr(>|t|)
iq0 110.019 4.305 25.556 < 2e-16 ***
m 77694.174 26708.502 2.909 0.00486 **
b 679.688 1039.144 0.654 0.51520
Residual standard error: 4054 on 70 degrees of freedom
> gwe <- lm(lynn2$rGDPpc ~ sf(lynn2$IQ,99.64)); summary(gwe)
Call:
lm(formula = lynn2$rGDPpc ~ sf(lynn2$IQ, 99.64))
Residuals:
Min 1Q Median 3Q Max
-10621.6 -2463.1 442.6 1743.4 12439.7
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -1345.7 886.6 -1.518 0.133
sf(lynn2$IQ, 99.64) 40552.7 2724.1 14.887 <2e-16 ***
Residual standard error: 4241 on 71 degrees of freedom
Multiple R-squared: 0.7574, Adjusted R-squared: 0.7539
F-statistic: 221.6 on 1 and 71 DF, p-value: < 2.2e-16
> opt <- lm(lynn2$rGDPpc ~ sf(lynn2$IQ,110)); summary(opt)
Call:
lm(formula = lynn2$rGDPpc ~ sf(lynn2$IQ, 110))
Residuals:
Min 1Q Median 3Q Max
-11030.3 -1540.1 -416.8 1308.6 12493.5
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 676.4 731.5 0.925 0.358
sf(lynn2$IQ, 110) 77577.0 4869.6 15.931 <2e-16 ***
Residual standard error: 4025 on 71 degrees of freedom
Multiple R-squared: 0.7814, Adjusted R-squared: 0.7783
F-statistic: 253.8 on 1 and 71 DF, p-value: < 2.2e-16
> his <- lm(lynn2$rGDPpc ~ sf(lynn2$IQ,108.2)); summary(his)
Call:
lm(formula = lynn2$rGDPpc ~ sf(lynn2$IQ, 108.2))
Residuals:
Min 1Q Median 3Q Max
-11014.0 -1710.0 -196.9 1396.3 12432.9
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 362.6 748.0 0.485 0.629
sf(lynn2$IQ, 108.2) 67711.5 4258.5 15.900 <2e-16 ***
Residual standard error: 4031 on 71 degrees of freedom
Multiple R-squared: 0.7807, Adjusted R-squared: 0.7777
F-statistic: 252.8 on 1 and 71 DF, p-value: < 2.2e-16
> em <- lm(log(lynn2$rGDPpc) ~ lynn2$IQ); summary(em)
Call:
lm(formula = log(lynn2$rGDPpc) ~ lynn2$IQ)
Residuals:
Min 1Q Median 3Q Max
-1.12157 -0.34268 -0.00503 0.29596 1.41540
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 0.812650 0.446308 1.821 0.0728 .
lynn2$IQ 0.089439 0.005051 17.706 <2e-16 ***
Residual standard error: 0.4961 on 71 degrees of freedom
Multiple R-squared: 0.8153, Adjusted R-squared: 0.8127
F-statistic: 313.5 on 1 and 71 DF, p-value: < 2.2e-16
Most importantly, we appear to have figured out the answer to my original question: no, it is not easy. :P
And inadvertently provided an object lesson for anyone watching about the value of researchers providing code...
The value of the smart fraction cutoff appears to have a huge impact on the napping from smart fraction to gdp, but doesn’t appear to have a significant effect on the goodness of fit, which troubles me somewhat. I’m also surprised that deleting the outliers seems to have improved the performance of the exponential fit more than the smart fraction fit, which is not what I would have expected from the graphs.
My intuition so far is that La Griffe found a convoluted way of regressing on a sigmoid, and the gain is coming from the part which looks like an exponential. I’m a little troubled that his stuff is so hard to reproduce sanely and that he doesn’t compare against the exponential fit: the exponent is obvious, has a reasonable empirical justification. Granting that Dickerson published in 2006 and he wrote the smart fraction essay in 2002 he could at least have updated.
[edit] Sorry, it looks like the formatting for my code is totally ugly.
You need to delete any trailing whitespace in your indented R terminal output. (Little known feature of LW/Reddit Markdown code blocks: one or more trailing spaces causes the newline to be ignored and the next line glommed on. I filed an R bug to fix some cases of it but I guess it doesn’t cover nls or you don’t have an updated version.)
I don’t understand your definition
egdp <- function(iq,iq0,m,b) (m*sf(iq,iq0)+b)
sf(iq,iq0) makes sense, of course, and m presumably is the multiplicative scale constant LG found to be 69k, but what is this b here and why is it being added? I don’t see how this tunes how big a smart fraction is necessary since shouldn’t it then be on the inside of sf somehow?
But using that formula and running your code (using the full dataset I posted originally, with outliers):
R> erf <- function(x) 2 * pnorm(x * sqrt(2)) - 1
R> sf <- function(iq,iq0) ((1+erf((iq-iq0)/(15*sqrt(2))))/2)
R> egdp <- function(iq,iq0,m,b) (m*sf(iq,iq0)+b)
R> summary(nls(rGDPpc ~ egdp(IQ,iq0,m,b), lynn, start=list(iq0=110,m=40000,b=0)))
Formula: rGDPpc ~ egdp(IQ, iq0, m, b)
Parameters:
Estimate Std. Error t value Pr(>|t|)
iq0 102.08 4.89 20.88 < 2e-16
m 37108.87 9107.73 4.07 0.00011
b 1140.94 1445.76 0.79 0.43241
Residual standard error: 5320 on 78 degrees of freedom
Number of iterations to convergence: 7
Achieved convergence tolerance: 5.09e-06
My intuition so far is that La Griffe found a convoluted way of regressing on a sigmoid, and the gain is coming from the part which looks like an exponential. I’m a little troubled that his stuff is so hard to reproduce sanely and that he doesn’t compare against the exponential fit: the exponent is obvious, has a reasonable empirical justification. Granting that Dickerson published in 2006 and he wrote the smart fraction essay in 2002 he could at least have updated.
I emailed La Griffe via Steve Sailer in February 2013 with a link to this thread and a question about how his smart-fraction model works with the fresher IQ/nations data and compares to Dickerson’s work. Sailer forwarded my email, but neither of us has had a reply since; he speculated that La Griffe may be having health issues.
In the absence of any defense by La Griffe, I think Dicker’s exponential works better than La Griffe’s fraction/sigmoid.
he doesn’t compare against the exponential fit: the exponent is obvious, has a reasonable empirical justification.
The theoretical justifications are entirely different, though. It seems reasonable to me to suppose there’s some minimal intelligence to be wealth-producing in an industrial society, and the smart fraction estimates that well and it predicts gdp well. But, it also seems reasonable to treat log(gdp) as a more meaningful object than gdp.
It’s also bothersome that the primary empirical prediction of the smart fraction model (that there is some stable gdp level that you hit when everyone is higher than the smart fraction) is entirely from the extrapolated part of the dataset, and this doesn’t seem noticeably better than the exponential model, whose extrapolations are radically different.
Granting that Dickerson published in 2006 and he wrote the smart fraction essay in 2002 he could at least have updated.
Yeah; I’m curious what they’d have to say about the relative merits of the two models. I’ll see if I can get this question to them.
You need to delete any trailing whitespace in your indented R terminal output.
Fixed, thanks!
but what is this b here and why is it being added?
It’s an offset, so that it’s an affine fit rather than a linear fit: the gdp level for a population with no people above 108 IQ doesn’t have to be 0. Turns out, it’s not significantly different from zero, but I’d rather discover that than enforce it (and enforcing it can degrade the value for m).
But, it also seems reasonable to treat log(gdp) as a more meaningful object than gdp.
I’m not entirely sure… For individuals, log-transforms make sense on their own merits as giving a better estimate of the utility of that money, but does that logic really apply to a whole country? More money means more can be spent on charity, shooting down asteroids, etc.
It’s also bothersome that the primary empirical prediction of the smart fraction model (that there is some stable gdp level that you hit when everyone is higher than the smart fraction) is entirely from the extrapolated part of the dataset, and this doesn’t seem noticeably better than the exponential model, whose extrapolations are radically different.
The next logical step would be to bring in the second 2006 edition of the Lynn dataset, which increased the set from 81 to 113, and use the latest available per-capita GDP (probably 2011). If the exponential fit gets better compared to the smart-fraction sigmoid, then that’s definitely evidence towards the conclusion that the smart-fraction is just a bad fit.
Yeah; I’m curious what they’d have to say about the relative merits of the two models. I’ll see if I can get this question to them.
I’d guess that he’d consider SF a fairly arbitrary model and not be surprised if an exponential fits better.
It’s an offset, so that it’s an affine fit rather than a linear fit: the gdp level for a population with no people above 108 IQ doesn’t have to be 0. Turns out, it’s not significantly different from zero, but I’d rather discover that than enforce it (and enforcing it can degrade the value for m).
Why can’t the GDP be 0 or negative? Afghanistan and North Korea are right now exhibiting what such a country looks like: they can barely feed themselves and export so much violence or fundamentalism or other dysfunctionality that rich nations are sinking substantial sums of money into supporting them and fixing problems.
For individuals, log-transforms make sense on their own merits as giving a better estimate of the utility of that money, but does that logic really apply to a whole country?
The argument would be that additional intelligence multiplies the per-capita wealth-producing apparatus that exists, rather than adding to it (or, in the smart fraction model, not doing anything once you clear a threshold).
Why can’t the GDP be 0 or negative?
There’s no restriction that b be positive, and so those are both options. I wouldn’t expect it to be negative because pre-industrial societies managed to survive, but that presumes that aid spending by the developed world is not subtracted from the GDP measurement of those countries. Once you take aid into account, then it does seem reasonable that places could become money pits.
The argument would be that additional intelligence multiplies the per-capita wealth-producing apparatus that exists, rather than adding to it (or, in the smart fraction model, not doing anything once you clear a threshold).
That’s the intuitive justification for an exponential model (each additional increment of intelligence adds a percentage of the previous GDP), but I don’t see how this justifies looking at log transforms.
There’s no restriction that b be positive, and so those are both options. I wouldn’t expect it to be negative because pre-industrial societies managed to survive
The difference would be a combination of negative externalities and changing Malthusian equilibriums: it has never been easier for an impoverished country like North Korea or Afghanistan to export violence and cause massive costs they don’t bear (9/11 directly cost the US something like a decade of Afghanistan GDP once you remove all the aid given to Afghanistan), and public health programs like vaccinations enable much larger populations than ‘should’ be there.
That’s the intuitive justification for an exponential model (each additional increment of intelligence adds a percentage of the previous GDP), but I don’t see how this justifies looking at log transforms.
GDP ~ exp(IQ) is isomorphic to ln(GDP) ~ IQ, and I think log(dollars per year) is an easier unit to think about than something to the power of IQ.
[edit] The graph might look different, though. It might be instructive to compare the two, but I think the relationships should be mostly the same.
It’s worth pointing out that IQ numbers are inherently non-parametric: we simply have a ranking of performance on IQ tests, which are then scaled to fit a normal distribution.
If GDP ~ exp(IQ), that means that the correlation is better if we scale the rankings to fit a log-normal distribution instead (this is not entirely true because exp(mean(IQ)) is not the same as mean(exp(IQ)), but the geometric mean and arithmetic mean should be highly correlated with each other as well). I suspect that this simply means that GDP approximately follows a log-normal distribution.
I suspect that this simply means that GDP approximately follows a log-normal distribution.
This doesn’t quite follow, since both per capita GDP and mean national IQ aren’t drawn from the same sort of distribution as individual production and individual IQ are, but I agree with the broader comment that it is natural to think of the economic component of intelligence measured in dollars per year as lognormally distributed.
...We find substantial impacts of salt iodization. High school completion rose by 6 percentage points, and labor force participation went up by 1 point. Analysis of income transitions by quantile shows that the new labor force joiners entered at the bottom of the wage distribution and took up blue collar labor, pulling down average wage income conditional on employment. Our results inform the ongoing debate on salt iodization in many low-income countries. We show that large-scale iodized salt distribution had a targeted impact, benefiting the worker on the margin of employment, and generating sizeable economic returns at low cost...The recent study by Feyrer et al. (2013) estimates that Morton Salt Co.’s decision to iodize may have increased IQ by 15 points, accounting for a significant part of the Flynn Effect, the steady rise IQ in the US over the twentieth century. Our estimates, paired with this number, suggest that each IQ point accounts for nearly one tenth of a point increase in labor force participation.
If, in the 1920s, 10 IQ points could increase your labor participation rate by 1%, then what on earth does the multiplier look like now? The 1920s weren’t really known for their demands on intelligence, after all.
And note the relevance to discussions of technological unemployment: since the gains are concentrated in the low end (think 80s, 90s) due to the threshold nature of iodine & IQ, this employment increase means that already, a century ago, people in the low-end range were having trouble being employed.
Social science research has shown that intelligence is positively correlated with patience and frugality, while growth theory predicts that more patient countries will save more. This implies that if nations differ in national average IQ, countries with higher average cognitive skills will tend to hold a greater share of the world’s tradable assets. I provide empirical evidence that in today’s world, countries whose residents currently have the highest average IQs have higher savings rates, higher ratios of net foreign assets to GDP, and higher ratios of U.S. Treasuries to GDP. These nations tend to be in East Asia and its offshoots. The relationship between national average IQ and net foreign assets has strengthened since the end of Bretton Woods.
...And time preference differs across countries in part because psychometric intelligence, a key predictor of patient behavior, differs persistently across countries (Wicherts et al., 2010a,b; Jones and Schneider, 2010)....John Rae (1834) provides a precursor of the approach presented here: Chapter Six of his treatise (cited in Becker and Mulligan, 1997, and Frederick et al., 2002) focuses on individual determinants of savings, including differences in rates of time preference, while his Chapter Seven draws out the cross-country implications...A recent meta-analysis of 24 studies by Shamosh and Gray concluded: “[A]cross studies, higher intelligence was associated with lower D[elay] D[iscounting]...” Their meta-study drew on experiments with preschool children and college students, drug addicts and relatively healthy populations: With few exceptions, they found a reliable relationship between measured intelligence and patience. And recent work by economists (Frederick, 2005; Benjamin et al. 2006; Burks et al, 2009; Chabris et al., 2007) has demonstrated that low-IQ individuals tend to act in a more “behavioral,” more impulsive fashion when facing decisions between smaller rewards sooner versus larger rewards later.
This is related, but not the research talked about. The Terman Project apparently found that the very highest IQ cohort had many more patents than the lower cohorts, but this did not show up as massively increased lifetime income.
Compare the bottom right IQ graph with SMPY results which show the impact of ability (SAT-M measured before age 13) on publication and patent rates. Ability in the SMPY graph varies between 99th and 99.99th percentile in quintiles Q1-Q5. The variation in IQ between the bottom and top deciles of the Terman study covers a similar range. The Terman super-smarties (i.e., +4 SD) only earned slightly more (say, 15-20% over a lifetime) than the ordinary smarties (i.e., +2.5 SD), but the probability of earning a patent (SMPY) went up by about 4x over the corresponding ability range.
Unless we want to assume those 4x extra patents were extremely worthless, or that the less smart groups were generating positive externalities in some other mechanism, this would seem to imply that the smartest were not capturing anywhere near the value they were creating—and hence were generating significant positive externalities.
EDIT: Jones 2011 argues much the same thing—economic returns to IQ are so low because so much of it is being lost to positive externalities.
On its own, I don’t consider this strong evidence for the greater productivity of the IQ elite. If they were contributions to open-source projects, that would be one thing. But people doing work that generates patents which don’t lead to higher income—that raises some questions for me. Is it possible that extremely high IQ is associated with a tendency to become “addicted” to a game like patenting? Added: I think Gwern and I agree more than many people might think reading this comment.
If they were contributions to open-source projects, that would be one thing.
Open-source contribution is even more gameable than patents: at least with patents there’s a human involved, checking to some degree that there is at least a little new stuff in the patent, while no one and nothing stops you from putting a worthless repo up on Github reinventing wheels poorly.
But people doing work that generates patents which don’t lead to higher income—that raises some questions for me.
The usual arrangement with, say, industrial researchers is that their employers receive the unpredictable dividends from the patents in exchange for forking over regular salaries in fallow periods...
Is it possible that extremely high IQ is associated with a tendency to become “addicted” to a game like patenting?
I don’t see why you would privilege this hypothesis.
Let me put it this way. Before considering the Terman data on patents you presented, I already thought IQ would be positively correlated with producing positive externalities and that there was a mostly one way causal link from the former to the latter. I expected the correlation between patents and IQ. What was new to me was the lack of correlation between IQ and income, and the lack of correlation between patents and income. Correction added: there was actually a fairly strong correlation between IQ and income, just not between income and patents, (conditional on IQ I think). Surely more productive industrial researchers are generally paid more. Many firms even give explicit bonuses on a per patent basis. So for me, given my priors, the Terman data you presented shifts me slightly against correction: does not shift me for or against the hypothesis that at the highest IQ levels, higher IQ individuals continues to be associated with producing more positive externalities. ref Still, I think increasing people’s IQ, even the already gifted, probably has strong positive externalities unless the method for increasing it also has surprising (to me) side-effects.
I agree that measuring open-source contributions requires more than merely counting lines of code written. But I did want to highlight the fact that the patent system is explicitly designed to increase the private returns for a given innovation. I don’t think that there is a strong correlation between the companies/industries which are patenting the most, and the companies/industries, which are benefiting the world the most.
Surely more productive industrial researchers are generally paid more. Many firms even give explicit bonuses on a per patent basis.
Yes, but the bonuses I’ve heard of are in the hundreds to thousands of dollars range, at companies committed to patenting like IBM. This isn’t going to make a big difference to lifetime incomes where the range is 1-3 million dollars although the data may be rich enough to spot these effects (and how many patents is even ‘4x’? 4 patents on average per person?), and I suspect these bonuses come at the expense of salaries & benefits. (I know that’s how I’d regard it as a manager: shifting risk from the company to the employee.)
And I think you’re forgetting that income did increase with each standard deviation by an amount somewhat comparable to my suggested numbers for patents, so we’re not explaining why IQ did not increase income whatsoever, but why it increased it relatively little, why the patenters apparently captured relatively little of the value.
Woh, I did allow myself to misread/misremember your initial comment a bit so I’ll dial it back slightly. The fact that even at the highest levels IQ is still positively correlated to income is important, and its what I would have expected, so the overall story does not undermine my support for the hypothesis that at the highest IQ levels, higher IQ individuals produce more positive externalities. I apologize for getting a bit sloppy there.
I would guess that if you had data from people with the same job description at the same company the correlation between IQ, patents, and income would be even higher.
Perhaps economic returns to IQ as so low because there are other skills which are good for getting economic returns, and those skills don’t correlate strongly with IQ.
Yes, this is consistent with the large income changes seen with some of the personality traits. If you have time, you could check the paper to see if that explains it: perhaps the highest cohort disproportionately went into academia or was low on Extraversion or something, or those subsets were entirely responsible for the excess patents.
they looked at (1) the economic gains to countries with higher average IQ, (2) the average gains to individuals with higher IQ, and concluded that (3) people with high IQ create vast amounts of positive externality, much more than they capture as individuals
How did they establish that economic gains are influenced by average IQ, rather than both being influenced by some other factor?
Sounds implausible to me, so I’m very interested in a citation (or pointers to similar material). If true, I’m going to have to do a lot of re-thinking.
Perhaps IQ correlates weakly with intelligence. If their are lots of people with high IQ, their are probably lots of intelligent people, but they’re not necessarily the same people. Hence, the countries with high IQ do well, but not the people.
It’s a nice list, but I think the core point strikes me as liable to be simply false. I forget who it was presenting this evidence—it might even have been James Miller, it was someone at the Winter Intelligence conference at FHI—but they looked at (1) the economic gains to countries with higher average IQ, (2) the average gains to individuals with higher IQ, and concluded that (3) people with high IQ create vast amounts of positive externality, much more than they capture as individuals, probably mostly in the form of countries with less stupid economic policies.
Maybe if we’re literally talking about a pure speed and LTM pill that doesn’t affect at all, say, capacity to keep things in short-term memory or the ability to maintain complex abstractions in working memory, i.e., a literal speed and disk space pill rather than an IQ pill.
Absolutely—IQ is very important, especially in aggregate. And yet, I’d still bet that the next day people will just be moving faster.
I think its worth making the distinction between having hardware which can support complex abstractions and actually having good decision making software in there. Although it’d be foolish to ignore the former because it tends to lead to the latter, it seems to be the latter that is more directly important.
That, and the fact that people can generally support better software than they pick up on their own is what makes our goal here doable.
If this is true, it would affect my decisions about whether and how to have children. So I’d really like to see the source if you can figure out what it was.
James Miller says:
That’s helpful; thanks.
Sounds plausible. If anybody finds the citation for this, please post it.
How about http://www.psychologicalscience.org/index.php/news/releases/are-the-wealthiest-countries-the-smartest-countries.html ?
Citing “Cognitive Capitalism: The impact of ability, mediated through science and economic freedom, on wealth”. (PDF not immediately available in Google.)
EDIT: efm found the PDF: http://www.tu-chemnitz.de/hsw/psychologie/professuren/entwpsy/team/rindermann/publikationen/11PsychScience.pdf
Or http://www.nickbostrom.com/papers/converging.pdf :
EDITEDIT: high IQ predicts superior stock market investing even after the obvious controls. High IQ types are also more likely to trust the stock market enough to participate more in it
“Do you have to be smart to be rich? The impact of IQ on wealth, income and financial distress”, Zagorsky 2007:
One could also phrase this as: “if we control for factors which we know to because by intelligence, such as highest level of education, then mirabile dictu! intelligence no longer increases income or wealth very much!”; or, “regressions are hard, let’s go shopping.”
Apropos of http://lemire.me/blog/archives/2012/07/18/why-we-make-up-jobs-out-of-thin-air/
Intelligence: A Unifying Construct for the Social Sciences, Lynn & Vanhanen 2012 (excerpts)
“IQ in the Ramsey Model: A Naïve Calibration”, Jones 2006:
That quote does not appear to come from the linked paper, and I’m confused as to how a paper from 2006 was supposed to have a citation from 2009.
Only the first paragraph is wrong (mixed it up with a paper on the Swiss iodization experience I’m using in a big writeup on iodide self-experimentation). Fixed.
“Economic gains resulting from the reduction in children’s exposure to lead in the United States”, Grosse et al 2002 (fulltext)
Their summary estimate from pg5/567 is a lower-middle-upperbound of each IQ point is worth, in net present value 2000 dollars: 12,700-14,500-17,200.
(Note that these figures, as usual, are net estimates of the value to an individual: so they are including zero-sum games and positional benefits. They aren’t giving estimates of the positive externalities or marginal benefits.)
“Quality of Institutions : Does Intelligence Matter?”, Kalonda-Kanyama & Kodila-Tedika 2012:
“IQ and Permanent Income: Sizing Up the “IQ Paradox””:
“Are Smarter Groups More Cooperative? Evidence from Prisoner’s Dilemma Experiments, 1959-2003”, Jones 2008:
Later: http://econlog.econlib.org/archives/2012/10/group_iq_one_so.html
What if higher SAT schools tend to be more prestigious and have stronger student identification?
Dunno. It’s consistent with all the other results about IQ and not school spirit...
Hm. Looks like going to a public/private school didn’t seem to mediate student cooperation all that much, which probably works against my theory.
They’re all US studies. Do we have anything from other cultures?
“IQ in the Production Function: Evidence from Immigrant Earnings”, Jones & Schneider 2008:
“Costs and benefits of iodine supplementation for pregnant women in a mildly to moderately iodine-deficient population: a modelling analysis” (mirror; appendices), Monahan et al 2015
IQ estimates:
All the details are in the Monahan et al 2015 appendices
The 8 studies are listed on pg8 of the appendix, Table 1:
Fletcher J. “Friends or Family? Revisiting the Effects of High School Popularity on Adult Earnings”. 2013. National Bureau of Economic Research Working Papers: 19232
Lutter RW. “Valuing children’s health: A reassessment of the benefits of lower lead levels”. AEI-Brookings Joint Center Working Paper No. 00-02. 2000.
Mueller G, Plug E. “Estimating the Effect of Personality on Male and Female Earnings”. Ind Lab Relat Rev. 2006;60(1):3-22.
Salkever DS. “Updated estimates of earnings benefits from reduced exposure of children to environmental lead”. Environ Res. 1995;70(1):1-6.
Schwartz J. “Societal benefits of reducing lead exposure”. Environ Res. 1994;66(1):105-24.
de Wolff P, van Slijpe ARD. “The Relation Between Income, Intelligence, Education and Social Background”. Europ Econ Rev. 1973;4(3):235-64.
Zax JS, Rees DI. IQ, “Academic Performance, Environment, and Earnings”. Rev Econ Stat. 2002;84(4):600-16
Zagorsky JL. “Do you have to be smart to be rich? The impact of IQ on wealth, income and financial distress”. Intelligence. 2007;35(5):489-501.
(Note that by including covariates that are obviously caused by IQ rather than independent, and excluding any attempt at measuring the many positive externalities of greater intelligence, these numbers can usually be considered substantial underestimates of country-wide benefits.)
“The High Cost of Low Educational Performance: the long-run economic impact of improving PISA outcomes”, Hanushek & Woessmann 2010:
Needless to say, “cognitive skills” here is essentially an euphemism for intelligence/IQ.
But but Goodhart’s law!
And it’s also confusing correlation with causation; grading is in large part due to intelligence. Boosting scores may be useless.
“Education, Intelligence, and Attitude Extremity”, Makowsky & Miller 2012
“The relationship between happiness and intelligent quotient: the contribution of socio-economic and clinical factors”, Ali et al 2012; effect is weakened once you take into account all the relevant variables but does sort of still exist.
I think that you might be confusing causation and correlation here. Countries that started to industrialize earlier have higher average IQ and higher GDP per capita. That would produce the effect you refer to. Whether or not the increased intelligence then contributes to further economic growth is a different matter.
What third factor producing both higher IQ and then industrialization are you suggesting?
Obviously you’re not suggesting anything as silly as the industrialization causes all observed IQ changes, because that simply doesn’t explain all examples, like East Asian countries:
That suggests that the correlation would have been less at that earlier time, which suggests the idea that the correlation of average IQ and average income has varied over history. Perhaps it has become stronger with increasing technological level—that is, more opportunities to apply smarts?
That certainly seems possible. Imagine a would-be programming genius who is born now, versus born in the Stone Age—he could become the wealthiest human to ever live (Bill Gates) or just the best hunter in the tribe (to be optimistic...).
Rindermann 2011: “Intellectual classes, technological progress and economic development: The rise of cognitive capitalism”; from abstract:
Here’s another one: “National IQ and National Productivity: The Hive Mind Across Asia”, Jones 2011
Above link is dead. Here is a new one
http://mason.gmu.edu/~gjonesb/JonesADR
“Exponential correlation of IQ and the wealth of nations”, Dickerson 2006:
It peeves me when scatterplots of GDP per capita versus something else use a linear scale—do they actually think the difference between $30k and $20k is anywhere near as important as that between $11k and $1k? And yet hardly anybody uses logarithmic scales.
Likewise, the fit looks a lot less scary if you write it as ln(GDP) = A + B*IQ.
Yes, Dickerson does point out that his exponential fit is a linear relationship on a log scale. For example, he does show a log-scale in figure 3 (pg3), fitting the most reliable 83 nation-points on a plot of log(GDP) against mean IQ in which the exponential fit looks exactly like you would expect. (Is it per capita? As far as I can tell, he always means per capita GDP even if he writes just ‘GDP’.) Figure 4 does the same thing but expands the dataset to 185 nations. The latter plot should probably be ignored given that the expansion comes from basically guessing:
Is it easy to compare the fit of their theory to the smart fraction theory?
I dunno. I’ve given it a try and while it’s easy enough to reproduce the exponential fit (and the generated regression line does fit the 81 nations very nicely), I think I screwed up somehow reproducing the smart fraction equation because the regression looks weird and trying out the smart-fraction function (using his specified constants) on specific IQs I don’t get the same results as in La Griffe’s table. And I can’t figure out what I’m doing wrong, my function looks like it’s doing the same thing as his. So I give up. Here is my code if you want to try to fix it:
(In retrospect, I’m not sure it’s even meaningful to try to fit the
sffunction with the constants already baked in, but since I apparently didn’t write it right, it doesn’t matter.Hm, one thing I notice is that you look like you’re fitting sf against log(gdp). I managed to replicate his results in octave, and got a meaningful result plotting smart fraction against gdp.
My guess at how to change your code (noting that I don’t know R):
That should give you some measure of how good it fits, and you might be able to loop it to see how well the smart fraction does with various thresholds.
(I also probably should have linked to the refinement.)
I can’t tell whether that works since you’re just using the same broken smart-fraction
sfpredictor; eg.sf(107,108)~> 32818, while the first smart fraction page’s table gives a Hong Kong regression line of 19817 which is very different from 33k.The refinement doesn’t help with my problem, no.
Hmmm. I agree that it doesn’t match. What if by ‘regression line’ he means the regression line put through the sf-gdp data?
That is, you should be able to calculate sf as a fraction with
And then regress that against gdp, which will give you the various coefficients, and a much more sensible graph. (You can compare those to the SFs he calculates in the refinement, but those are with verbal IQ, which might require finding that dataset / trusting his, and have a separate IQ0.)
Comparing the two graphs, I find it interesting that the eight outliers Griffe mentions (Qatar, South Africa, Barbados, China, and then the NE Asian countries) are much more noticeable on the SF graph than the log(GDP) graph, and that the log(GDP) graph compresses the variation of the high-income countries, and gets most of its variation from the low-income countries; the situation is reversed in the SF graph. Since both our IQ and GDP estimates are better in high-income countries, that seems like a desirable property to have.
With outliers included, I’m getting R=.79 for SF and R=.74 for log(gdp). (I think, I’m not sure I’m calculating those correctly.)
Trying to rederive the constants doesn’t help me, which is starting to make me wonder if he’s really using the table he provided or misstated an equation or something:
If you double 34779 you get very close to his $69,321 so there might be something going wrong due to the 1⁄2 that appears in uses of the
erfto make a cumulative distribution function, but I don’t how a threshold of 99.64 IQ is even close to his 108!(The weird start values were found via trial-and-error in trying to avoid R’s ‘singular gradient error’; it doesn’t appear to make a difference if you start with, say,
f=90.)Most importantly, we appear to have figured out the answer to my original question: no, it is not easy. :P
So, I started off by deleting the eight outliers to make lynn2. I got an adjusted R^2 of 0.8127 for the exponential fit, and 0.7777 for the fit with iq0=108.2.
My nls came back with an optimal iq0 of 110, which is closer to the 108 I was expecting; the adjusted R^2 only increases to 0.7783, which is a minimal improvement, and still slightly worse than the exponential fit.
The value of the smart fraction cutoff appears to have a huge impact on the mapping from smart fraction to gdp, but doesn’t appear to have a significant effect on the goodness of fit, which troubles me somewhat. I’m also surprised that deleting the outliers seems to have improved the performance of the exponential fit more than the smart fraction fit, which is not what I would have expected from the graphs. (Though, I haven’t calculated this with the outliers included in R, and I also excluded the Asian data, and there’s more fiddling I can do, but I’m happy with this for now.)
And inadvertently provided an object lesson for anyone watching about the value of researchers providing code...
My intuition so far is that La Griffe found a convoluted way of regressing on a sigmoid, and the gain is coming from the part which looks like an exponential. I’m a little troubled that his stuff is so hard to reproduce sanely and that he doesn’t compare against the exponential fit: the exponent is obvious, has a reasonable empirical justification. Granting that Dickerson published in 2006 and he wrote the smart fraction essay in 2002 he could at least have updated.
You need to delete any trailing whitespace in your indented R terminal output. (Little known feature of LW/Reddit Markdown code blocks: one or more trailing spaces causes the newline to be ignored and the next line glommed on. I filed an R bug to fix some cases of it but I guess it doesn’t cover
nlsor you don’t have an updated version.)I don’t understand your definition
sf(iq,iq0)makes sense, of course, andmpresumably is the multiplicative scale constant LG found to be 69k, but what is thisbhere and why is it being added? I don’t see how this tunes how big a smart fraction is necessary since shouldn’t it then be on the inside ofsfsomehow?But using that formula and running your code (using the full dataset I posted originally, with outliers):
I emailed La Griffe via Steve Sailer in February 2013 with a link to this thread and a question about how his smart-fraction model works with the fresher IQ/nations data and compares to Dickerson’s work. Sailer forwarded my email, but neither of us has had a reply since; he speculated that La Griffe may be having health issues.
In the absence of any defense by La Griffe, I think Dicker’s exponential works better than La Griffe’s fraction/sigmoid.
The theoretical justifications are entirely different, though. It seems reasonable to me to suppose there’s some minimal intelligence to be wealth-producing in an industrial society, and the smart fraction estimates that well and it predicts gdp well. But, it also seems reasonable to treat log(gdp) as a more meaningful object than gdp.
It’s also bothersome that the primary empirical prediction of the smart fraction model (that there is some stable gdp level that you hit when everyone is higher than the smart fraction) is entirely from the extrapolated part of the dataset, and this doesn’t seem noticeably better than the exponential model, whose extrapolations are radically different.
Yeah; I’m curious what they’d have to say about the relative merits of the two models. I’ll see if I can get this question to them.
Fixed, thanks!
It’s an offset, so that it’s an affine fit rather than a linear fit: the gdp level for a population with no people above 108 IQ doesn’t have to be 0. Turns out, it’s not significantly different from zero, but I’d rather discover that than enforce it (and enforcing it can degrade the value for m).
I’m not entirely sure… For individuals, log-transforms make sense on their own merits as giving a better estimate of the utility of that money, but does that logic really apply to a whole country? More money means more can be spent on charity, shooting down asteroids, etc.
The next logical step would be to bring in the second 2006 edition of the Lynn dataset, which increased the set from 81 to 113, and use the latest available per-capita GDP (probably 2011). If the exponential fit gets better compared to the smart-fraction sigmoid, then that’s definitely evidence towards the conclusion that the smart-fraction is just a bad fit.
I’d guess that he’d consider SF a fairly arbitrary model and not be surprised if an exponential fits better.
Why can’t the GDP be 0 or negative? Afghanistan and North Korea are right now exhibiting what such a country looks like: they can barely feed themselves and export so much violence or fundamentalism or other dysfunctionality that rich nations are sinking substantial sums of money into supporting them and fixing problems.
The argument would be that additional intelligence multiplies the per-capita wealth-producing apparatus that exists, rather than adding to it (or, in the smart fraction model, not doing anything once you clear a threshold).
There’s no restriction that b be positive, and so those are both options. I wouldn’t expect it to be negative because pre-industrial societies managed to survive, but that presumes that aid spending by the developed world is not subtracted from the GDP measurement of those countries. Once you take aid into account, then it does seem reasonable that places could become money pits.
That’s the intuitive justification for an exponential model (each additional increment of intelligence adds a percentage of the previous GDP), but I don’t see how this justifies looking at log transforms.
The difference would be a combination of negative externalities and changing Malthusian equilibriums: it has never been easier for an impoverished country like North Korea or Afghanistan to export violence and cause massive costs they don’t bear (9/11 directly cost the US something like a decade of Afghanistan GDP once you remove all the aid given to Afghanistan), and public health programs like vaccinations enable much larger populations than ‘should’ be there.
GDP ~ exp(IQ) is isomorphic to ln(GDP) ~ IQ, and I think log(dollars per year) is an easier unit to think about than something to the power of IQ.
[edit] The graph might look different, though. It might be instructive to compare the two, but I think the relationships should be mostly the same.
It’s worth pointing out that IQ numbers are inherently non-parametric: we simply have a ranking of performance on IQ tests, which are then scaled to fit a normal distribution.
If GDP ~ exp(IQ), that means that the correlation is better if we scale the rankings to fit a log-normal distribution instead (this is not entirely true because exp(mean(IQ)) is not the same as mean(exp(IQ)), but the geometric mean and arithmetic mean should be highly correlated with each other as well). I suspect that this simply means that GDP approximately follows a log-normal distribution.
This doesn’t quite follow, since both per capita GDP and mean national IQ aren’t drawn from the same sort of distribution as individual production and individual IQ are, but I agree with the broader comment that it is natural to think of the economic component of intelligence measured in dollars per year as lognormally distributed.
“Salt Iodization and the Enfranchisement of the American Worker”, Adhvaryu et al 2013:
If, in the 1920s, 10 IQ points could increase your labor participation rate by 1%, then what on earth does the multiplier look like now? The 1920s weren’t really known for their demands on intelligence, after all.
And note the relevance to discussions of technological unemployment: since the gains are concentrated in the low end (think 80s, 90s) due to the threshold nature of iodine & IQ, this employment increase means that already, a century ago, people in the low-end range were having trouble being employed.
A 2012 Jones followup: “Will the intelligent inherit the earth? IQ and time preference in the global economy”
This is related, but not the research talked about. The Terman Project apparently found that the very highest IQ cohort had many more patents than the lower cohorts, but this did not show up as massively increased lifetime income.
http://infoproc.blogspot.com/2011/04/earnings-effects-of-personality.html
Unless we want to assume those 4x extra patents were extremely worthless, or that the less smart groups were generating positive externalities in some other mechanism, this would seem to imply that the smartest were not capturing anywhere near the value they were creating—and hence were generating significant positive externalities.
EDIT: Jones 2011 argues much the same thing—economic returns to IQ are so low because so much of it is being lost to positive externalities.
On its own, I don’t consider this strong evidence for the greater productivity of the IQ elite. If they were contributions to open-source projects, that would be one thing. But people doing work that generates patents which don’t lead to higher income—that raises some questions for me. Is it possible that extremely high IQ is associated with a tendency to become “addicted” to a game like patenting? Added: I think Gwern and I agree more than many people might think reading this comment.
Open-source contribution is even more gameable than patents: at least with patents there’s a human involved, checking to some degree that there is at least a little new stuff in the patent, while no one and nothing stops you from putting a worthless repo up on Github reinventing wheels poorly.
The usual arrangement with, say, industrial researchers is that their employers receive the unpredictable dividends from the patents in exchange for forking over regular salaries in fallow periods...
I don’t see why you would privilege this hypothesis.
Let me put it this way. Before considering the Terman data on patents you presented, I already thought IQ would be positively correlated with producing positive externalities and that there was a mostly one way causal link from the former to the latter. I expected the correlation between patents and IQ. What was new to me was the lack of correlation between IQ and income, and the lack of correlation between patents and income. Correction added: there was actually a fairly strong correlation between IQ and income, just not between income and patents, (conditional on IQ I think). Surely more productive industrial researchers are generally paid more. Many firms even give explicit bonuses on a per patent basis. So for me, given my priors, the Terman data you presented shifts me slightly against correction: does not shift me for or against the hypothesis that at the highest IQ levels, higher IQ individuals continues to be associated with producing more positive externalities. ref Still, I think increasing people’s IQ, even the already gifted, probably has strong positive externalities unless the method for increasing it also has surprising (to me) side-effects.
I agree that measuring open-source contributions requires more than merely counting lines of code written. But I did want to highlight the fact that the patent system is explicitly designed to increase the private returns for a given innovation. I don’t think that there is a strong correlation between the companies/industries which are patenting the most, and the companies/industries, which are benefiting the world the most.
Yes, but the bonuses I’ve heard of are in the hundreds to thousands of dollars range, at companies committed to patenting like IBM. This isn’t going to make a big difference to lifetime incomes where the range is 1-3 million dollars although the data may be rich enough to spot these effects (and how many patents is even ‘4x’? 4 patents on average per person?), and I suspect these bonuses come at the expense of salaries & benefits. (I know that’s how I’d regard it as a manager: shifting risk from the company to the employee.)
And I think you’re forgetting that income did increase with each standard deviation by an amount somewhat comparable to my suggested numbers for patents, so we’re not explaining why IQ did not increase income whatsoever, but why it increased it relatively little, why the patenters apparently captured relatively little of the value.
Woh, I did allow myself to misread/misremember your initial comment a bit so I’ll dial it back slightly. The fact that even at the highest levels IQ is still positively correlated to income is important, and its what I would have expected, so the overall story does not undermine my support for the hypothesis that at the highest IQ levels, higher IQ individuals produce more positive externalities. I apologize for getting a bit sloppy there.
I would guess that if you had data from people with the same job description at the same company the correlation between IQ, patents, and income would be even higher.
Perhaps economic returns to IQ as so low because there are other skills which are good for getting economic returns, and those skills don’t correlate strongly with IQ.
Yes, this is consistent with the large income changes seen with some of the personality traits. If you have time, you could check the paper to see if that explains it: perhaps the highest cohort disproportionately went into academia or was low on Extraversion or something, or those subsets were entirely responsible for the excess patents.
3 more links:
“The Role of Cognitive Skills in Economic Development”
“The Role of School Improvement in Economic Development”
“An economic and rational choice approach to the autism spectrum and human neurodiversity” (Tyler Cowen)
If anyone is curious, I am moving my bibliography here to http://www.gwern.net/Embryo%20selection#value-of-iq and I will be keeping that updated in the future rather than continue this thread further.
How did they establish that economic gains are influenced by average IQ, rather than both being influenced by some other factor?
Sounds implausible to me, so I’m very interested in a citation (or pointers to similar material). If true, I’m going to have to do a lot of re-thinking.
Perhaps IQ correlates weakly with intelligence. If their are lots of people with high IQ, their are probably lots of intelligent people, but they’re not necessarily the same people. Hence, the countries with high IQ do well, but not the people.
I think you really need to see this google tech talk by Steven Hsu.